一种基于压缩感知的大规模多入多出系统信道压缩反馈方法

A Unique Massive Multiple Input Multiple Output Channel Feedback Based on Compressed Sensing

大规模无线多入多出系统由于在无线收发端配置了多天线构成的天线阵列,信道数量大幅度增加,传统的信道反馈方法面临巨大挑战。鉴此,提出一种基于压缩感知的信道信息压缩反馈方案,以减少计算复杂度与降低反馈量。结合双向主成分分析理论推导在空间相关瑞利信道下的双向投影矩阵,可以有效地将信道信息矩阵稀疏化,并给出了一个合适的比特分配方案。与主成分分析压缩信道算法对比,本方法可明显提升信道反馈的归一化均方误差(NMSE)性能,降低计算复杂度。

Due to the antenna array composed of multiple antennas,the number of channels in the massive multiple input multiple output(MIMO)system increases greatly,and the traditional channel feedback method faces great challenges.In view of this,a channel state information compression feedback scheme based on compressed sensing is proposed to reduce computational complexity and feedback overhead.Based on the bidirectional principal component analysis theory,the projection matrix is derived for spatially correlated Rayleigh channels.Compared with the principal component analysis compressed channel algorithm,this method can significantly improves the NMSE performance of channel feedback and reduces the computational complexity and feedback overhead.